Non-Destructive Evaluation (NDE) of Composite Joints for Through Life Condition Monitoring in the Marine Environment

Abstract

A key barrier to the use of composite superstructures on naval ships remains proving the integrity of adhesively bonded composite to steel joints. Hence, the work examines existing Non-Destructive Evaluation (NDE) methods, exploring opportunities to adapt them for maritime applications. Novel inspection technologies are developed alongside tools that allow for full field comparisons to numerical data for the validation of numerical models.

An investigation of existing NDE techniques identified Pulse Thermography (PT) as a technique with potential for adaptation for maritime applications. While well-established for the inspection of thin laminates and adhesive joints, errors inherent to PT inspections can obscure the presence of defects in thicker laminates and joints. Hence, a novel processing routine was developed, improving probing depth by compensating for errors in both the temporal and frequency domains. The result was a 200% increase in probing depth allowing for the identification and characterisation of defects at depths not possible using existing PT techniques. Whilst extending the applicability of PT to new application areas, this improvement is not sufficient to inspect typical maritime applications, where laminate thicknesses often exceed 6 mm.

Therefore, Lock-in Thermography (LIT) was explored as an alternative. This existing technique was significantly adapted to form a low cost inspection system based on internal heating. For this purpose a novel embedded actuator is demonstrated provide internal heating for LIT inspections. Low cost micro-bolometer infrared (IR) cameras are combined with the development of a simple modulation control circuit, reducing inspection equipment costs by an order of magnitude whilst maintaining performance traditionally associated with photon detector based IR cameras (approx. 1.8 mm probing depth). The proposed system increases the accessibility of thermography to new researchers and industrial applications, facilitating new research and increased industrial uptake of thermography.

The embedded material concept is further exploited, and its functionality exploited as a novel electrically conductive embedded sacrificial sensor, capable of detecting damage in composite materials and bonded joints. The sensor is demonstrated in Single Lap Joints (SLJ), where the sensor electrical response is sensitive to damage onset. Interlaminar shear tests and shear lap failure loads of SLJs show no reduction in laminate or joint strength when sensors are embedded. A high resolution Digital Image Correlation (DIC) setup confirms damage initiation and provides a new perspective of the effect of spew fillets configurations in SLJs. Possible secondary uses are identified for load estimation and load cycle counting.

With a method established for damage detection in thick composite laminates, the later part of the thesis focusses on fusing full-field experimental data with high fidelity numerical models of adhesively bonded joints. It is demonstrated that full-field pointwise comparisons can be made between DIC data and strain data obtained from models a identical resolution achieved using data interpolation. This work forms an important first step in developing high fidelity models which could be used to assess the criticality of damage identified in NDE inspection. Such models could prove invaluable post damage identification, where decisions on whether to impose operation restrictions, maintenance planning and remedial work are based on quantified data.

Overall, the work extends the applicability of existing NDE techniques, by improving data processing. A new low cost thermography approach is developed that reduces costs by at least a factor of ten and opens up possibilities of using thermography for continuous monitoring. A new embedded sensor is conceived to effectively identify damage within an adhesive joint. Important initial steps are taken towards developing tool for data comparison based on integrating high fidelity numerical models with full-field experimental data.

More information

Identifiers

Catalogue record

Export record

Altmetrics